Circuit for automatic selection between speech and music sound signals

Abstract

Speech versus music is decided by comparing the presence of pauses greater than 32, 60 and 500 ms in the lower input spectrum, and 4 ms in the higher input spectrum.

We claim: 
     
       1. Electric circuit for automatic switching of a sound reproduction and signal apparatus for exercising a selected choice, among selectable signals, between speech and music sound reproduction signals, comprising: means (40) for registering said selected choice;   means (12,13) for subjecting a selectable or selected sound reproduction signal, to be tested for conformance with a choice registered in said registering means, to separation of said signal into a low-frequency signal component and a high-frequency signal component, said low-frequency signal component containing substantially all component frequencies of said sound reproduction signals which are less than 3 kHz;   means (14-19) for converting said low-frequency and high-frequency signal components respectively into first and second sequences of amplitude-limited pulses;   first timing means (24) responsive to said first sequence of amplitude-limited pulses for detecting pauses in said low-frequency signal component which pauses exceed a first predetermined duration;   second timing means (25) for identifying periods of time in which said pauses detected by said first timing means succeed each other at intervals not exceeding a second predetermined duration;   third timing means (26) responsive to said second sequence of amplitude-limited pulses for detecting the occurrence of pauses in said high-frequency signal component exceeding a third predetermined duration which is less than said first predetermined duration;   fourth timing means (23) responsive to said first sequence of amplitude-limited pulses for detecting the occurrence of pauses in said low-frequency signal component exceeding a fourth predetermined duration which is longer than said third predetermined duration and shorter than said first predetermined duration;   first logic circuit means (27,27a) responsive to the outputs (a,d) of said first and fourth timing means for detecting the absence of a pause exceeding said third predetermined duration in said high-frequency signal component during the portion of a pause in said low-frequency signal component which exceeds said fourth predetermined duration and thereby producing a first logic output signal (e) and for detecting the absence of a pause exceeding said third predetermined duration in said high-frequency signal component during the absence of pauses exceeding said fourth predetermined duration in said low-frequency signal component and thereby producing a second logic output signal (f);   second logic circuit means (31) for dientifying a music signal by a third logic output signal (h) with reference to the respective states of said first and second logic output signals (e,f) and the state of the output (c) of said second timing means (25) and for identifying a speech signal by a fourth logic signal (i) with a reference to the respective states of said first and third logic signals (e,h) and the respective states of the outputs (b,c) of said first and second timing means (24,25), and   sound reproduction signal evaluating means (33-39) for comparing the state of said registering means (40) with the corresponding one of said third and fourth logic signals and determining in response thereto whether or not the sound reproduction signal selected for test is of the character selected by said registering means.   
     
     
       2. Circuit as defined in claim 1, in which said second predetermined duration is of a magnitude in the range between 400 and 2000 ms. 
     
     
       3. Circuit as defined in claim 1, in which said first predetermined duration is of a magnitude in the range between 50 and 200 ms. 
     
     
       4. Circuit as defined in claim 1 in which differentiating means (20,21) are interposed between the respective outputs of said converting means and the inputs of said timing means respectively responsive to said first and second sequences of amplitude limited pulses. 
     
     
       5. Circuit as defined in claim 4 in which a noise-blanker (22) is interposed between said means for converting said low-frequency signal component into said first series of amplitude-limited pulses and said differentiating means (20) in which a signal inverter (29) is connected to the output of said noise-blanker (22) to provide a supplementary speech-indicating-factor signal (g) at the output of said inverter, and in which said second logic circuit means (31) are constituted to identify a speech signal with reference not only to the respective states of said first and third logic signals (e,h) and of the outputs (b,c) of said first and second timing means, but also with reference to said supplementary speech-indicating-factor signal (g). 
     
     
       6. Circuit as defined in claim 5, in which means (39) are provided for selecting measurement timing intervals between measurement clock pulses of which the intervals are of a magnitude suitable for speech or music signal identification as selected according to the state of said registering means (40) designating a selection between speech and music, and in which latches are provided for storing output signals of said first logic circuit means (27,27a) and said first and second timing means (24,25) during a measuring period and in which said second logic circuit means (31) are constituted to identify a music signal in response to the absence of pauses exceeding said first predetermined duration in said low-frequency signal component, the absence of pauses exceeding said third predetermined duration in said high-frequency signal component during the absence of pauses exceeding said fourth predetermined duration in said low-frequency signal component, and the occurrence of a pause in said high-frequency signal component during the portion of every pause in said low-frequency component which exceeds said fourth predetermined duration. 
     
     
       7. Circuit as defined in claim 1, in which said first predetermined duration is approximately 60 ms. 
     
     
       8. Circuit as defined in claim 6, in which said second logic circuit means (31) are constituted for identifying a speech signal when within a measuring period, at least the following conditions have been detected: the absence of identification of a music signal; the presence of a pause in said low-frequency signal component which has exceeded said first predetermined duration; the presence of a quiescent condition of the output of said noise blanker, and either the absence of occurrence of a pause exceeding said third predetermined duration in said high-frequency signal component during the portion exceeding said fourth predetermined duration of a pause in said low-frequency signal component or else the continued existence of an interval less than said second predetermined duration since the last previous time when a pause in said low-frequency signal component began to exceed said first predetermined duration. 
     
     
       9. Circuit as defined in claim 7, in which said third predetermined duration is of a magnitude in the range between 3 and 50 ms. 
     
     
       10. Circuit as defined in claim 9, in which said fourth predetermined duration is of a magnitude in the range between 10 and 50 ms. 
     
     
       11. Circuit as defined in claim 8, in which said sound reproduction signal evaluating means include control logic (37) means and also means for storing said third logic output signals or said fourth logic output signals produced in successive measurement periods, following successive resettings of said signal storage latches, and determining only on the basis of three in succession out of five successive measurement periods, each following a resetting of the latches, a final identification of speech or music signal content, said control logic means (37) being further constituted so as to initiate a search for another signal source if the operation of said sound reproduction signal evaluating means does not confirm a choice registered in said registering means said third of said fourth logic output signals being selected for storage according to said choice registered in said registering means. 
     
     
       12. Circuit as defined in claim 11, in which a counter is provided in said sound reproduction signal evaluating means for counting said measurement periods and thereby counting out a measurement cycle of five measurement periods. 
     
     
       13. Circuit as defined in claim 1, in which means (39) are provided for selecting measurement timing intervals between measurement clock pulses of which the intervals are of a magnitude suitable for speech or music signal identification as selected according to the state of said registering means (40) designating a selection between speech and music, and in which latches are provided for storing output signals of said first logic circuit means (27,27a) and said first and second timing means (24,25) during a measuring period and in which said second logic circuit means (31) are constituted to identify a music signal in response to the absence of pauses exceeding said first predetermined duration in said low-frequency signal component, the absence of pauses exceeding said third predetermined duration in said high-frequency signal component during the absence of pauses exceeding said fourth predetermined duration in said low-frequency signal component, and the occurrence of a pause in said high-frequency signal component during a portion of every pause in said low-frequency component which exceeds said fourth predetermined duration. 
     
     
       14. Circuit as defined in claim 1 in which said second logic circuit means (31) are constituted for taking account, in the identification of a speech signal, of the state (g) of said first sequence of amplitude-limited pulses. 
     
     
       15. Circuit as defined in claim 1 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 kHz. 
     
     
       16. Circuit as defined in claim 13 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 kHz. 
     
     
       17. Circuit as defined in claim 4 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 kHz. 
     
     
       18. Circuit as defined in claim 24 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 kHz. 
     
     
       19. Circuit as defined in claim 5 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 KHz. 
     
     
       20. Circuit as defined in claim 6 in which said signal component separating means are constituted to produce said high-frequency signal component in a form in which said high-frequency signal component contains substantially no component frequencies below 6 kHz and contains substantially all the component frequencies of said sound reproduction signal lying between 6 kHz and 10 kHz.